CN1642220A - Image processing device, image display device, image processing method, and image processing program - Google Patents

Abstract

Aspects of the invention relate to a method of processing images capable of performing natural saturation enhancement having a balance between color components with a small amount of calculation is provided. The image processing device may be arranged in an image display device or may be a device for providing display image data to the external image display device to perform the saturation enhancement of the input image data. Specifically, saturation information can be extracted from the input image data to determine a saturation enhancement amount. In addition, a color component to which the saturation enhancement is performed can be extracted from the input image data. Further, based on the determined saturation enhancement amount, the saturation enhancement for each color component can be performed. Here, for the predetermined color component, the saturation is enhanced after the saturation enhancement amount is adjusted. The color component included in the input image data may be, for example, Cb and Cr in the YCbCr color space. The saturation enhancement amount is adjusted to the predetermined color component. Therefore, the color balance of the image data after the saturation can be properly adjusted and the natural image data can be obtained.

Description

Image processing apparatus, method, program and image display device

Technical field

The present invention relates to the image processing method of the colourity of correcting colour images.

Background technology

Among image display device etc., when the colourity of input color image is not enough, emphasize the treatment for correcting of colourity.Recording and narrating in patent documentation 1 has this colourity to emphasize the example of handling.The method of patent documentation 1 is at first to detect maximum chrominance from the coloured image of input, generates the chromaticity correction table with maximum chrominance as the upper limit.At this moment, the chromaticity correction table is set at colourity with output image data curve more than or equal to the colourity of input image data.So, use this chromaticity correction table to carry out chromaticity correction and handle.In this document, because generate the chromaticity correction table after detecting maximum chrominance at first, the output behind the chromaticity correction can not surpass maximum chrominance, has the advantage that so-called " color collapse " can not occur.Yet the method for patent documentation 1 is owing to all must exist the huge problem of calculation process amount by input color image data computation maximum chrominance at every turn.

In addition, in colourity is emphasized to handle, be mostly in the color space of RGB, YUV generally, any colouring component is all emphasized equally., when any colouring component was all emphasized equally, the balance between colouring component can be damaged, and it is unnatural that the image after handling sometimes becomes.

Patent documentation 1: the spy opens the 2001-218078 communique.

Summary of the invention

The invention that the present invention system finishes in view of the problems referred to above, its problem is to provide a kind of can carry out the image processing method that the balance between the colouring component is emphasized for very natural colourity with few amount of calculation.

In a viewpoint of the present invention, image processing apparatus has: extract the chrominance information of input image data and determine the colourity amount of emphasizing of the colourity amount of emphasizing to determine the unit; From above-mentioned input image data, extract the colouring component extraction unit of a plurality of colouring components of wanting the Accent colour degree; According to the colourity amount of emphasizing of decision above-mentioned a plurality of colouring components are carried out the colourity that colourity emphasizes and emphasize the unit; And among above-mentioned a plurality of colouring components, in the colourity of predetermined colouring component is emphasized, the amount of the emphasizing adjustment unit that the colourity amount of emphasizing of above-mentioned decision is adjusted.

Above-mentioned image processing apparatus can be arranged on the inside of image display device, or as device formation from display image data to the image display device of outside that supply with, carry out the colourity of input image data and emphasize.In specific words, be the decision colourity amount of emphasizing when from input image data, extracting chrominance information.In addition, also from input image data extraction the carrying out colouring component that colourity is emphasized.So,, carry out the colourity of each colouring component is emphasized according to the colourity amount of emphasizing of decision.Herein, the colouring component for predetermined after the adjustment of carrying out the colourity amount of emphasizing, carries out colourity and emphasizes.As the colouring component that is contained in the input image data, such as, can enumerate Cb in the YCbCr color space and Cr etc.For predetermined colouring component, by adjustment to the colourity amount of emphasizing, the colour balance of the view data after can suitably adjusting colourity and emphasizing and obtain the view data of nature.

A kind of mode of above-mentioned image processing apparatus, the luminance information extraction unit with luminance information of extracting above-mentioned input image data, the above-mentioned amount of emphasizing adjustment unit is adjusted the above-mentioned colourity amount of emphasizing according to the luminance information of extracting.Carrying out colourity when emphasizing, in specific colouring component, the briliancy according to input image data difference occurs in the emphasis effect of colourity sometimes.Such as, in the Cb component, when briliancy is high, in the blue component suitable, produce strangulation (clip) and the colourity emphasis effect is reduced with the positive direction of Cb component, owing to the yellow color component that is equivalent to negative direction is not produced the colourity emphasis effect that strangulation obtains expecting, so the emphasizing of yellow color component compared too much with blue component, can produce destruction of balance.Therefore, by predetermined colouring component is carried out the adjustment of the colourity amount of emphasizing according to luminance information, can prevent above-mentioned defective.

In a kind of mode of above-mentioned image processing apparatus, the above-mentioned amount of emphasizing adjustment unit when the briliancy that is comprised in the luminance information of said extracted is high, reduces the above-mentioned colourity amount of emphasizing.In the example of above-mentioned Cb component etc., when briliancy is high, reduce by the negative direction colourity amount of emphasizing that makes the Cb component, the positive direction of Cb component and the amount of emphasizing of negative direction (being blue component and yellow color component) are equated, can keep colour balance.

In a kind of mode of above-mentioned image processing apparatus, the above-mentioned amount of emphasizing adjustment unit is littler than the amount of emphasizing in the positive region by the amount of adjusting in the negative region that the above-mentioned colourity amount of emphasizing can make the above-mentioned predetermined colouring component in the color space of emphasizing.Thus, can keep the colour balance of the view data after emphasizing.

In a preferred embodiment, the above-mentioned amount of emphasizing adjustment unit makes factor alpha (0＜α＜1) that the amount of emphasizing in the negative region of above-mentioned predetermined colouring component becomes the amount of emphasizing in the positive region doubly.Herein, preferably can set above-mentioned factor alpha is fixed value at the mean value of above-mentioned luminance information during smaller or equal to predetermined value, then becomes less than the said fixing value greater than predetermined value along with the mean value of above-mentioned luminance information becomes.

In addition, in another preferred embodiment, above-mentioned image processing apparatus has the briliancy correcting unit of the briliancy component that is contained in the above-mentioned input image data being proofreaied and correct according to the luminance information of extracting, and above-mentioned factor alpha is and the irrelevant fixed value of above-mentioned luminance information.Along with chromaticity correction carries out the briliancy timing simultaneously, because it is many when on the direction that briliancy increases, proofreading and correct, such as, reduce without exception by making the colourity amount of emphasizing with the corresponding colouring component of the negative direction of Cb component (being yellow color component), can keep colour balance.

In a preferred embodiment, the above-mentioned predetermined colouring component of the above-mentioned color space is the Cb colouring component in the YCbCr color space.

In another preferred embodiment of above-mentioned image processing apparatus, above-mentioned input image data is the animation data that is made of the multiple image data, has the scene change-detection unit that generates the degree control signal corresponding that the scene with above-mentioned animation data changes, the above-mentioned amount of emphasizing adjustment unit, by at the corresponding above-mentioned factor alpha of above-mentioned multiple image, implement and the corresponding weighting of above-mentioned control signal, can calculate the above-mentioned factor alpha of this two field picture.In view of the above, can set suitable factor alpha by the front and back that change in scene, the colourity that just can obtain nature is emphasized.

Have above-mentioned image processing apparatus and show the image display device of display part of the output of above-mentioned image processing apparatus by formation, can show under the situation that the good colourity of colour balance emphasizes in that input image data is carried out.

In another viewpoint of the present invention, image processing method has: extract the chrominance information of input image data and determine the colourity amount of the emphasizing deciding step of the colourity amount of emphasizing; From above-mentioned input image data, extract the colouring component extraction step of a plurality of colouring components of wanting the Accent colour degree; And according to the colourity amount of emphasizing of above-mentioned decision above-mentioned a plurality of colouring components are carried out the colourity that colourity emphasizes and emphasize step; Among above-mentioned a plurality of colouring components, in the colourity of predetermined colouring component is emphasized, the colourity amount of emphasizing of above-mentioned decision is adjusted.Utilize this image processing method, also the same with the above-mentioned image processing apparatus colourity that can suitably keep colour balance is emphasized.

In the viewpoint of another one more of the present invention, image processing program by carrying out on computers, can make the aforementioned calculation facility that the chrominance information of extracting input image data is arranged and determines the colourity amount of the emphasizing decision unit of the colourity amount of emphasizing; From above-mentioned input image data, extract the colouring component extraction unit of a plurality of colouring components of wanting the Accent colour degree; According to the colourity amount of emphasizing of decision above-mentioned a plurality of colouring components are carried out the colourity that colourity emphasizes and emphasize the unit; And among above-mentioned a plurality of colouring components, in the colourity of predetermined colouring component is emphasized, the function of the amount of the emphasizing adjustment unit that the colourity amount of emphasizing of above-mentioned decision is adjusted.By carrying out this image processing program on computers, can realize above-mentioned image processing apparatus.

Description of drawings

Fig. 1 illustrates the distribution of the pixel value in the color space.

Fig. 2 illustrates the distribution of the pixel value in the Y-Cb plane.

Fig. 3 illustrates the distribution of the pixel value in the Y-Cr plane.

Fig. 4 illustrates the variation of the pixel value of chromaticity transformation curve.

Fig. 5 illustrates the variation of the frequency distribution of chromaticity transformation curve.

Fig. 6 illustrates the distribution of the pixel value of the big image of brightness value.

The frequency that Fig. 7 illustrates the pixel value when strangulation occurring distributes.

Fig. 8 illustrates the example of the conversion curve of the amount of emphasizing that suppresses the Cb component.

The frequency of the pixel value when Fig. 9 illustrates the amount of emphasizing that suppresses the Cb component distributes.

Figure 10 illustrates the setting example of adjusting parameter.

Figure 11 is the block diagram that the summary of image display device of the present invention constitutes.

Figure 12 is the flow chart that the tint correction of embodiment 1 is handled.

Figure 13 is the diagrammatic sketch of explanation sample area.

Figure 14 is the curve of frequency distribution illustrated example of briliancy.

Figure 15 is the curve chart that the calculated example of the colourity amount of emphasizing is shown.

Figure 16 is the example of chromaticity correction table.

Figure 17 illustrates the setting example of the adjustment parameter of embodiment 1.

Figure 18 is the flow chart that the tint correction of embodiment 2 is handled.

Figure 19 is the calculated example of briliancy correcting value.

Figure 20 is the setting example of briliancy check point.

Figure 21 proofreaies and correct with the example of showing for briliancy.

Figure 22 illustrates the setting example of the adjustment parameter of embodiment 2.

Figure 23 is the flow chart that the tint correction of embodiment 3 is handled.

Symbol description

10 image processing apparatus 11CPU 12ROM 13RAM, 20 display parts, 100 image display devices

Embodiment

Before preferred embodiment is described, at first colourity of the present invention is emphasized that the summary of handling is illustrated.

Fig. 1 illustrates the distribution situation of pixel value in the color space of input picture.As the color space, what use is the YCbCr color space herein.Y represents briliancy (brightness) component, and Cb, Cr represent colouring component.

Because Cb, Cr are the components of expression look, the big expression of its absolute value is bright-colored.Fig. 2, Fig. 3 represent from the Y-Cb plane respectively, the distribution of the pixel value of Y-Cr viewed in plan.Any one figure represents that because colourity is emphasized, the scope of pixel value is according to arrow range expansion laterally.At this moment, because the absolute value of Cb, Cr is big, can increase look vividness (promptly carrying out colourity emphasizes).

In fact, utilize the conversion curve of Fig. 4 to carry out emphasizing of colourity.Conversion curve is the approximate S font by initial point, and transverse axis is input, and the longitudinal axis is output.When utilizing this conversion curve to carry out the pixel value conversion of Cb, Cr, can make absolute value become big.

The variation of absolute value can be expressed as the variation that frequency shown in Figure 5 distributes.The transverse axis of Fig. 5 illustrates the pixel value of Cb (or Cr), and the longitudinal axis is the frequency of this pixel value.Utilize the conversion curve absolute value to become big, its result, frequency is distributed in both sides and broadens.In addition, because colourity emphasizes to make frequency to distribute when changing, total pixel number is constant.Because this frequency distribution table shows colouring component and distribute, when frequency distributed expansion as shown in Figure 5, that just means that carrying out colourity emphasizes.

Fig. 6 illustrates the distribution from the pixel value of Y-Cb viewed in plan at another input picture.This input picture is characterised in that brightness value Y is than higher.When carrying out for this image, pixel value can occur and reach phenomenon outside the color space with previous same conversion.In fact owing to can not get the outer value of the color space, get the value (this phenomenon of getting the value on border is called in generation " strangulation " (shearing) of the border of the color space) on color space border.

The frequency that Fig. 7 illustrates when strangulation occurring distributes.Known to from the figure on Y-Cb plane shown in Figure 2, the YCbCr color space in the high zone of briliancy, narrows down on the positive direction of Cb, broadens on the negative direction of Cb.Therefore, as shown in Figure 6, the strangulation pilosity is born in the positive region of Cb, and the expansion that the frequency of the positive region of Cb distributes can be more than or equal to certain value.On the other hand, the negative region of Cb, because the expansion of the color space is difficult to take place strangulation, it is big that the expansion that frequency distributes becomes.

In addition, known to from the figure on Y-Cr plane shown in Figure 3, have the YCbCr space, in the high zone of briliancy, on the positive direction of Cr, narrow down, the tendency that on the negative direction of Cr, broadens, the occasion of this and Cb is similar.But, more as can be known, in the high zone of briliancy, the positive region of the Cr in YCbCr space is narrow unlike the positive region of Cb with Fig. 2 and Fig. 3.So above-mentioned strangulation can not produce the problem as the Y-Cb space in the Y-Cr space.

As mentioned above, particularly in the high occasion of the briliancy of input picture, because strangulation creates a difference in the frequency expansion.This colourity that becomes relevant each colouring component is emphasized the difference of occasion, and the balance collapse when emphasizing with colourity is related.The positive direction of Cb is with blue corresponding, and the negative direction of Cb is with yellow corresponding.Therefore, in specific words, when the colourity that only relates to the negative direction of Cb is strongly emphasized, its result will take place with the corresponding yellow direction of the negative direction of Cb on the problem of collapse of colour balance.

To because the color space expansion of the negative direction of Cb and when relating to colourity strongly and emphasizing to be noted, as the main points that address the above problem, can make Cb negative direction emphasize to be weaker than other directions.The example of the conversion curve of the amount of emphasizing of the negative direction of inhibition Cb shown in Figure 8.In Fig. 8, set this coefficient of α (0＜α＜1) (adjustment parameter), the amount of emphasizing of the negative direction of Cb is compared with the general value A of the amount of emphasizing of the positive direction of Cb, can be suppressed.

The conversion curve of the Fig. 8 of utilization shown in Figure 9 carries out the frequency of colourity when emphasizing and distributes.As previously mentioned, in the positive direction of Cb, because strangulation, the expansion of distribution can be more than or equal to certain value.On the other hand, on the negative direction of Cb, also as shown in Figure 8, owing to import the amount of emphasizing that factor alpha has suppressed conversion curve, it is big that the expansion of distribution can not become.Its result, colourity emphasize that the horizontality that relates to is identical on the positive direction of Cb and negative direction, can prevent the collapse of colour balance.

In addition, as shown in Figure 6, particularly in the big image of briliancy, because the strangulation of the positive direction of Cb takes place, the factor alpha that had before illustrated (0＜α＜1) preferably makes the brightness value interlock of itself and input picture.In fact, because the briliancy of input picture has distribution,, use the mean value of briliancy as typical value.

Figure 10 illustrates the mean value that transverse axis is a briliancy, and the longitudinal axis is the diagrammatic sketch of setting example of the factor alpha of factor alpha.By setting, when making mean value in briliancy high, emphasize that quantitative change is little to factor alpha.

Like this, in the present invention, owing to the reason of the shape in YCbCr space, consider in the high zone of briliancy, in the positive direction generation strangulation of Cb,, adjust the colour balance on the positive negative direction of Cb and make the colourity of nature emphasize to become possibility by inhibition to the amount of emphasizing on the negative direction of Cb

[embodiment 1]

Below the embodiment that uses the image display device that chromaticity correction of the present invention handles is illustrated.In Figure 11 (a), the summary that image display device 100 is shown constitutes.Image display device 100 is to handling from the input image data of outside and device shown, the adducible AV of having equipment and have the terminal installation etc. of image display function.In addition, chromaticity correction processing of the present invention has operand and just can carry out this feature less, considers from this point, be specially adapted to, such as, portable phone and PDA (personal digital assistant) etc. are subjected to the mobile terminal device of certain limitation on operational capability.

Image display device 100 has to be implemented to comprise the image processing apparatus 10 of the image processing that chromaticity correction handles and shows the display part 20 of the image after handling input image data.Image processing apparatus 10, such as, shown in Figure 11 (b), its formation comprises the ROM12 of CPU11, memory image handling procedure and as RAM13 of working storage or the like.At this moment, chromaticity correction of the present invention is handled, and image processing program is stored in the ROM12 in advance, and CPU11 reads it and carry out and as image processing apparatus.In addition, also can not to carry out the image processing comprise chromaticity correction by software processes like this, but in image processing apparatus 10, carry out its part or all by hardware handles.

Display part 20, because of image display device 100 different, such as, can be liquid crystal display screen and other flat-panel monitors etc.

Below the chromaticity correction of embodiment 1 is handled and illustrated.In embodiment 1, as previously mentioned,, carry out the good chromaticity correction of colour balance by the amount of emphasizing in the negative region that is suppressed at the Cb component in the YCbCr space.

Figure 12 illustrates the flow chart of the chromaticity correction processing of embodiment 1.The step that the chromaticity correction of embodiment 1 is handled is as follows: the statistic of calculating input image data, extract input image data feature, generate checking list and utilize this checking list to carry out the chromaticity correction of input image data according to the feature of extracting.

In specific words, at first, input image data is sampled, generate curve of frequency distribution figure (step S1).Generally, provide because input image data is a frame image data as RGB, image processing apparatus 10 (comprising R, G, B) is transformed to the input image data of RGB form the view data (comprising Y, Cb, Cr) of YCbCr form according to formula (1) and (2).

[formula 1]

$\left(\begin{array}{c}Y\\ \mathrm{Cb}\\ \mathrm{Cr}\end{array}\right)=M\left(\begin{array}{c}R\\ G\\ B\end{array}\right)---\left(1\right)$

[formula 2]

$M=\left(\begin{array}{ccc}0.2990& 0.5870& 0.1140\\ 0.1687& 0.3313& 0.5000\\ 0.5000& -0.4187& -0.0813\end{array}\right)---\left(2\right)$

Formula (1) is the transform that is transformed to the view data of YCbCr form from the view data of RGB form, and formula (2) is its transform matrix M.

When obtaining the view data of YCbCr form, image processing apparatus 10 is sampled to it and is extracted the pixel generation curve of frequency distribution figure that is contained in the two field picture.The example that becomes the sample area of the object that utilizes sampling to extract pixel is shown in Figure 13 (a) and reaches (b).The example of Figure 13 (a) is with the example of entire frame image-region as sample area.When the size in length and breadth of two field picture was respectively Rx, Ry, total frequency of sampling was whole pixel count Rx * Ry.

The example of Figure 13 (b) is with the center of the two field picture example as sample area.When the size in length and breadth of two field picture is respectively Rx, Ry, when the ratio of the area of the relative entire image in center is R, the Ra of size in length and breadth, the Rb of center, such as, can utilize formula (3) to calculate.

[formula 3]

$R_a=\sqrt{R}{R}_x$ $R_b=\sqrt{R}{R}_y---\left(3\right)$

In addition, the coordinate of the summit A～D in the termination district shown in Figure 13 (b) is provided by formula (4).

[formula 4]

$A:(\frac{R_x-R_a}{2},\frac{R_y-R_b}{2})---B:(\frac{R_x+R_a}{2},\frac{R_y-R_b}{2})$

$C:(\frac{R_x-R_a}{2},\frac{R_y+R_b}{2})---D:(\frac{R_x+R_a}{2},\frac{R_y+R_b}{2})---\left(4\right)$

The zone can generate the curve of frequency distribution f that represents briliancy Y thus _A(k) and the curve of frequency distribution f of colourity S _S(k) formula (5).

[formula 5]

f _A(k)，k＝0，...，255

f _S(k)，k＝0，...，128 (5)

Wherein, colourity S utilizes formula (6) to calculate from color difference signal Cb, Cr.

[formula 6]

S＝{|C _b|+|C _r|}/2 (6)

The curve of frequency distribution of the briliancy Y that draws like this is illustrated in Figure 14.Transverse axis is tonal gradation number (0～255), and the longitudinal axis is frequency (pixel count with each gray value).In addition, generate the curve of frequency distribution figure of colourity S too.

Afterwards, utilize the curve of frequency distribution figure that obtains, compute statistics (step S2).In specific words, according to formula (7), respectively from the curve of frequency distribution f of briliancy Y _A(k) calculating mean value P _AaveReach curve of frequency distribution f from colourity S _S(k) calculating mean value P _Save

[formula 7]

$P_{\mathrm{Aave}}=\underset{k=0}{\overset{255}{\mathrm{\Σ}}}{\mathrm{kf}}_A\left(k\right)/\left(R_a{R}_b\right)$

$P_{\mathrm{Save}}=\underset{k=0}{\overset{128}{\mathrm{\Σ}}}{\mathrm{kf}}_S\left(k\right)/\left(R_a{R}_b\right)---\left(7\right)$

Afterwards, calculate the correction parameter (step S3) of chromaticity correction according to resulting statistic.In specific words, utilize the colourity average value P _Save, calculate the colourity amount of emphasizing P according to formula (8) _Sdiff'.

[formula 8]

${P_{\mathrm{Sdiff}}}^{\′}=\left\{\begin{array}{c}{P}_{\mathrm{Sdiff}},P_{\mathrm{Sdiff}}\≤P_{\mathrm{Sth}}\\ P_{\mathrm{Sth}},P_{\mathrm{Sdff}}>P_{\mathrm{Sth}}\end{array}\right.$

$P_{\mathrm{Sdiff}}=\left\{\begin{array}{c}(P_{\mathrm{Saveth}}-P_{\mathrm{Save}})P_{\mathrm{Scvl}}/P_{\mathrm{Sev}2},P_{\mathrm{Save}}<{\mathrm{<figure-callout\; id="0"\; label="P\; Saveth"\; filenames="A20051000059900242.png,A20051000059900301.png"\; state="\{\{state\}\}">P</figure-callout>}}_{\mathrm{Saveth}}\\ 0,P_{\mathrm{Save}}\&GreaterEqual;P_{\mathrm{Saveth}}\end{array}\right.---\left(8\right)$

Wherein, P _SavethBe the fiducial value of proofreading and correct, P _Scv1And P _Scv2Be the parameter of adjusting correcting value, P _SthIt is the limit of restriction correcting value.The colourity average value P _SaveWith the resulting colourity amount of emphasizing P _Sdiff' relation be shown in Figure 15.As can be seen from Figure 15, in the colourity average value P _SaveHour, make the colourity amount of emphasizing P _Sdiff' strengthen, with the degree of staining average value P _SaveIncrease, the colourity amount of emphasizing P _Sdiff' diminish gradually, in the colourity average value P _SaveWhen surpassing predetermined value, the colourity amount of emphasizing P _Sdiff' become 0 (promptly not emphasizing).Thus, can emphasize the part of colourity deficiency.

When obtaining correction parameter like this, according to correction parameter, calculation correction table (step S4).Checking list S[Cr at Cr component shown in Figure 16 (a)] example.Checking list S[Cr], generate by (0,0), (128 ,-128), (128,128), check point 1 (P _Spt, P _Spt'), check point 2 (P _Spt,-P _Spt') curve.When formation curve, use suitable interpolating functions such as spline interpolation.The former relatively point symmetry of check point 1 and 2, checking list S[Cr] also former relatively point symmetry.The output P of check point _Spt', can utilize input P _SptWith the colourity amount of emphasizing P _Sdiff' obtain according to formula (9).

[formula 9]

P _Spt′＝P _Spt+P _Sdiff′

P _Spt″＝P _Spt+P _Sdiff′·α _s (9)

At the checking list C[Cr shown in Figure 16 (a)] in, in the positive region and negative region of input data, use the same colourity amount of emphasizing P _Sdiff'.But, at the same checking list S[Cb that constitutes colouring component Cb] time, as mentioned above, when the briliancy of input image data was high, during strangulation, emphasizing in the negative region of Cb component strengthened the Cb component comparatively speaking, and the such defective of colour balance collapse takes place in positive region.Therefore, in the present invention, for checking list S[Cb], in negative region, import and adjust parameter alpha, the colourity amount of emphasizing is decided to be P _Sdiff' α _SThis moment checking list S[Cb] example be shown in Figure 16 (b).In addition, the output P of Ci Shi check point _Spt", can utilize input P _Spt, the colourity amount of emphasizing P _Sdiff' and adjust parameter alpha _SObtain by formula (9).Because adjustment parameter alpha _SBe 0＜α _S＜1, shown in Figure 16 (b), littler than the amount of emphasizing of positive region according to the amount of emphasizing of the negative region of the Cb component of checking list.In view of the above, the emphasizing of the negative region of Cb component, promptly yellow emphasizes, with other colouring components emphasize compare, be suppressed, can prevent the collapse of colour balance.

Afterwards, utilize the checking list S[Cb that generates in the above described manner] and S[Cr], carry out tint correction (step S5).Tint correction is carried out each pixel of input image data.In specific words, use checking list S[Cb] and S[Cr], generate view data Y ', Cb ' and Cr ' after proofreading and correct according to formula (10), be transformed to view data R ', G ' and the B ' of RGB form and export.In addition, in the present embodiment, briliancy Y does not proofread and correct.Like this, finishing chromaticity correction handles.

[formula 10]

$\begin{array}{c}{Y}^{\&prime;}=Y\\ C{b}^{\&prime;}=S_1\left[\mathrm{Cb}\right]\\ {\mathrm{Cr}}^{\&prime;}=S_2\left[\mathrm{Cr}\right]\end{array}\left(\begin{array}{c}{R}^{\&prime;}\\ G^{\&prime;}\\ B^{\&prime;}\end{array}\right)=M^{-1}\left(\begin{array}{c}{Y}^{\&prime;}\\ {\mathrm{Cb}}^{\&prime;}\\ {\mathrm{Cr}}^{\&prime;}\end{array}\right)---\left(10\right)$

Below to adjusting parameter alpha _SValue describe.α shown in Figure 17 _SAn example of curve chart.Transverse axis is the briliancy average value P _Aave, from the briliancy average value P _AaveThe ground that surpasses predetermined value is lighted and is made the adjustment parameter alpha _SValue decay.Thus, when the briliancy of input image data is high, suppress the degree of emphasizing in the negative region of Cb component, can prevent yellow emphasize excessive.α shown in Figure 17 _SValue can provide by formula (11).In addition, in the example of Figure 17, be from the briliancy average value P _AaveThe place that surpasses predetermined value makes the adjustment parameter alpha _SValue linear attenuation (dullness reduces), decay but also can replace with curve mode (with exponential function form).

[formula 11]

${\mathrm{\&alpha;}}_s=\left\{\begin{array}{c}1.0,P_{\mathrm{Aave}}<P_{\mathrm{Aave\&alpha;}}\\ a_{\mathrm{\&alpha;}}{P}_{\mathrm{Aave}}+b_{\mathrm{\&alpha;}},P_{\mathrm{Aave}}\&GreaterEqual;P_{\mathrm{Aave\&alpha;}}\end{array}\right\}$

$a_{\mathrm{\&alpha;}}=\frac{{\mathrm{\&alpha;}}_{\mathrm{th}}-1}{255-P_{\mathrm{Aave\&alpha;}}}$ b _α＝1.0-a _αP _Aaveα (11)

[embodiment 2]

Below embodiment 2 is illustrated.In embodiment 1, handle the occasion of only colourity being proofreaied and correct as tint correction and be illustrated, and in embodiment 2,, also briliancy is also proofreaied and correct simultaneously except colourity.But the bearing calibration of colourity is identical with embodiment 1.

In Figure 18, the flow chart of the tint correction processing of embodiment 2 is shown.The same with embodiment 1, at first, input image data is sampled, generate curve of frequency distribution figure (step S11) to briliancy Y and colourity S, calculate the statistic (step S12) of briliancy Y and colourity S according to this.

Afterwards, according to statistic calculation correction parameter (step S13).Herein, the correction parameter P of colourity S _Sdiff' with embodiment 1 the same being shown among Figure 15, can draw by formula (8).On the other hand, the correction parameter of briliancy Y (gamma correction parameter) P _Gdiff' as shown in figure 19, can draw by formula (12).In specific words, Figure 19 illustrates average value P _Aave' and correction parameter P _Gdiff' relation.

[formula 12]

${P_{\mathrm{Gdiff}}}^{\&prime;}=\left\{\begin{array}{c}{P}_{\mathrm{Gth}1},P_{\mathrm{Gdiff}}<P_{\mathrm{Gth}1}\\ P_{\mathrm{Gdiff}},P_{\mathrm{Gth}1}\&le;P_{\mathrm{Gdiff}}\&le;P_{\mathrm{Gth}2}\\ P_{\mathrm{Gth}2},P_{\mathrm{Gth}2}<P_{\mathrm{Gdiff}}\end{array}\right.---\left(12\right)$

P _Gdiff＝(P _Aaveth-P _Aave′)P _Gcv1/P _Gcv2

Wherein, P _GavethBe the fiducial value of proofreading and correct, P _Gcv1And P _Gcv2Be the parameter of adjusting correcting value, P _Gth1And P _Gth2It is the limit of restriction correcting value.Correction reference value P _Gaveth, in average value P _Aave' when being worth, applying the correction that makes the briliancy deepening, and when being worth, apply the correction that briliancy is brightened less than this greater than this.

Afterwards, according to the correction parameter of briliancy Y and colourity S, calculation correction table (step S14).With the embodiment 1 the same checking list that generates colourity S.On the other hand, the check point P of briliancy Y _GptCan obtain by formula (13), as shown in figure 20.

[formula 13]

$P_{\mathrm{Gpt}}=\left\{\begin{array}{c}{P}_{\mathrm{Gpt}2},{P_{\mathrm{Aave}}}^{\&prime;}<P_{\mathrm{Aaveth}}\\ P_{\mathrm{Gpt}1},{P_{\mathrm{Aave}}}^{\&prime;}\&GreaterEqual;P_{\mathrm{Aaveth}}\end{array}\right.---\left(13\right)$

By formula (13), in average value P _Aave' than correction reference value P _GavethWhen big, i.e. during briliancy deepening, the setting check point diminishes and (becomes P _Gpt1), otherwise when brightening, setting the check point change (becomes P greatly _Gpt2).Utilize this check point,, can obtain checking list (gamma checking list) Y of briliancy Y shown in Figure 21 by formula (14) _Gamma[Y].

[formula 14]

P _Gpt′＝P _Gpt+P _Gdiff′ (14)

Afterwards, utilize the checking list of briliancy Y and colourity S, carry out tint correction (step S15).In specific words, according to the briliancy Y ' behind formula (15) calculation correction and aberration Cb ' and Cr ', it is transformed to view data R ', G ' and the B ' of RGB form and exports.

[formula 15]

$\begin{array}{c}{Y}^{\&prime;}=Y_{\mathrm{gamma}}\left[Y\right]\\ {\mathrm{Cb}}^{\&prime;}=S_1\left[\mathrm{Cb}\right]\\ {\mathrm{Cr}}^{\&prime;}=S_2\left[\mathrm{Cr}\right]\end{array}\left(\begin{array}{c}{R}^{\&prime;}\\ G^{\&prime;}\\ B^{\&prime;}\end{array}\right)=M^{-1}\left(\begin{array}{c}{Y}^{\&prime;}\\ {\mathrm{Cb}}^{\&prime;}\\ {\mathrm{Cr}}^{\&prime;}\end{array}\right)---\left(15\right)$

Like this, end comprises the tint correction of the treatment for correcting of briliancy and colourity.

The adjustment parameter alpha of the chromaticity correction table of embodiment 2 _SExample provide and be shown in Figure 21,22 by formula (16).

[formula 16]

α _S＝α _th (16)

In embodiment 1, as shown in figure 17, adjusting parameter alpha _SMore than or equal to predetermined briliancy average value P _AaveThe time it is decayed gradually.Relative therewith, in the preferred embodiment of embodiment 2, as shown in figure 22, adjust parameter alpha _SBe fixed as predetermined value _ThIn embodiment 2, be not colourity, briliancy is also proofreaied and correct.Herein, in the briliancy average value P _Aave' hour, when promptly image is dark, carry out correction towards bright image.In Figure 17 of embodiment 1, in the briliancy average value P _Aave' hour, strangulation is difficult to take place, and makes the adjustment parameter alpha _SBe fixed as 1, and in embodiment 2, owing to make the correction of dark image towards bright image by the briliancy correction, because chromaticity correction, the possibility that strangulation takes place colouring component Cb increases on positive direction.So, in the embodiment 2 of the correction of carrying out briliancy Y, by making the adjustment parameter alpha _SBe fixed in predetermined value _Th(0＜α _S＜1), more to the colouring component (being yellow color component) of the negative region of colouring component Cb emphasize suppress, it is preferred adjusting colour balance.In addition, as another example, shown in the dotted line C1 of Figure 22, more than or equal to predetermined briliancy average value P _AaveThe zone in, make the adjustment parameter alpha _SFrom predetermined value _ThRise to increase gradually and also can.This be because, such as, in the briliancy average value P _Aave' big, when promptly image is bright, proofread and correct towards dark image, its result, because chromaticity correction, the possibility that strangulation takes place on positive direction colouring component Cb reduces event.

[embodiment 3]

Below embodiment 3 is illustrated.In embodiment 3, be when input image data is animation, carry out the detection of scene change and upgrade the adjustment parameter alpha _SGenerally when animation data being carried out briliancy correction and chromaticity correction, change when big in scene (scene), use the correction parameter (dynamically updating correction parameter) of this time point, on the other hand, change when little in scene, for preventing flicker, use same correction parameter (correction parameter does not dynamically update) for preferred.Therefore, for the adjustment parameter alpha of above-mentioned chromaticity correction _S, it is preferred upgrading with the form of the testing result that changes corresponding to scene.In specific words, get the average weighted while, average weighted weight is controlled according to the testing result that scene changes in adjustment parameter to continuous animation frame.

Now, such as, the adjustment parameter of two field picture is α before supposing _S0, the adjustment parameter of current frame image is α _S1, get average weighted adjustment parameter alpha _SCan illustrate by following formula:

α＝W _pre·α _so+W _now·α _S1 (17)

At this, W _PreAnd W _NowBe respectively for before with current frame image in the weight of parameter, get W _Pre+ W _Now=1.

Herein, such as, switching when big in scene, weight setting is W _Pre=0, W _Now=1.By setting this weight, the adjustment parameter alpha after the weighted average calculation of being determined by formula (17) becomes the adjustment parameter alpha of current frame image _S1, be updated to the adjustment parameter of this time point.Its result can carry out and follow the optimum correction that scene changes.

On the other hand, change when little in scene, weight setting is W _Pre=W _Now=0.5.By setting this weight, the adjustment parameter alpha after the weighted average calculation of being determined by formula (17) becomes simple average, can remove and adjust parameter small variations in time.Its result can alleviate the obstruction of flicker after the correction etc.In addition, illustrative herein is to detect to have or not scene to switch, but also can be to calculate the degree that scene changes, and according to this intensity of variation, average weighted weight is carried out continuous control.

Like this, when input image data is animation, change, adopt the weighted average of adjusting parameter alpha with the corresponding colourity of this scene change-detection result,, also can carry out suitable chromaticity correction even when input image data is animation by detecting scene.

Figure 23 illustrates the flow chart of the tint correction processing of embodiment 3.In addition, this embodiment is the same with embodiment 1, and what illustrate is to handle the occasion of only colourity being proofreaied and correct as tint correction, but as embodiment 2, and in the occasion that briliancy and colourity are proofreaied and correct, embodiment 3 also can use.

In Figure 23, step S21～S23 is identical with step S1～S3 of embodiment 1, each two field picture to input image data (being animation in this example) is sampled, generate curve of frequency distribution (step S21), compute statistics (step S22), calculation correction parameter (step S23).

Processing (step S24) when afterwards, carrying out the scene variation.That is, detect scene and change, calculate above-mentioned weighting W according to scene change-detection result _Pre, W _NowSo, calculate the adjustment parameter alpha according to formula (17).In addition, the scene change-detection such as, the difference of the briliancy mean value etc. of two field picture before and after can calculating, during more than or equal to predetermined value, just judges that occurrence scene changes in this difference.In addition, the whole bag of tricks of the detection method that changes as scene is known, describes in detail and omits.

Like this, during the adjustment parameter alpha upgraded obtaining accepting scene change-detection result, be according to generating checking list (step S25), carrying out tint correction (step S26) according to the checking list that generates with it.

Claims (12)

1. image processing apparatus is characterized in that comprising:

Extract the colourity amount of the emphasizing decision unit of the chrominance information decision colourity amount of emphasizing of input image data;

From above-mentioned input image data, extract the colouring component extraction unit of a plurality of colouring components of wanting the Accent colour degree;

According to the colourity amount of emphasizing of decision above-mentioned a plurality of colouring components are carried out the colourity that colourity emphasizes and emphasize the unit; And

Among above-mentioned a plurality of colouring components, in the colourity of predetermined colouring component is emphasized, the amount of the emphasizing adjustment unit that the colourity amount of emphasizing of above-mentioned decision is adjusted.

2. image processing apparatus as claimed in claim 1 is characterized in that:

Luminance information extraction unit with luminance information of extracting above-mentioned input image data;

The above-mentioned amount of emphasizing adjustment unit is adjusted the above-mentioned colourity amount of emphasizing according to the luminance information of extracting.

3. image processing apparatus as claimed in claim 2 is characterized in that:

The above-mentioned amount of emphasizing adjustment unit when the briliancy that is comprised in the luminance information of said extracted is high, reduces the above-mentioned colourity amount of emphasizing.

4. image processing apparatus as claimed in claim 2 is characterized in that:

The above-mentioned amount of emphasizing adjustment unit, the amount of adjusting in the negative region that the above-mentioned colourity amount of emphasizing makes the above-mentioned predetermined colouring component in the color space of emphasizing is littler than the amount of emphasizing in the positive region.

5. image processing apparatus as claimed in claim 4 is characterized in that:

The above-mentioned amount of emphasizing adjustment unit makes factor alpha (0＜α＜1) that the amount of emphasizing in the negative region of above-mentioned predetermined colouring component becomes the amount of emphasizing in the positive region doubly.

6. image processing apparatus as claimed in claim 5 is characterized in that:

Above-mentioned factor alpha is fixed value at the mean value of above-mentioned luminance information during smaller or equal to predetermined value, then becomes less than the said fixing value greater than predetermined value along with the mean value of above-mentioned luminance information becomes.

7. image processing apparatus as claimed in claim 5 is characterized in that:

Has the briliancy correcting unit of the briliancy component that is contained in the above-mentioned input image data being proofreaied and correct according to the luminance information of extracting;

Above-mentioned factor alpha is and the irrelevant fixed value of above-mentioned luminance information.

8. image processing apparatus as claimed in claim 4 is characterized in that:

The above-mentioned predetermined colouring component of the above-mentioned color space is the Cb colouring component in the YCbCr color space.

9. image processing apparatus as claimed in claim 5 is characterized in that:

Above-mentioned input image data is the animation data that is made of the multiple image data;

Has the scene change-detection unit that generates the degree control signal corresponding that the scene with above-mentioned animation data changes;

The above-mentioned amount of emphasizing adjustment unit, by at the corresponding above-mentioned factor alpha of above-mentioned multiple image, implement and the corresponding weighting of above-mentioned control signal, calculate the above-mentioned factor alpha of this two field picture.

10. image display device is characterized in that comprising:

Aforesaid right requires 1 to 9 any one described image processing apparatus; And

Demonstration is from the display part of the output of above-mentioned image processing apparatus.

11. an image processing method is characterized in that comprising:

Extract the colourity amount of the emphasizing deciding step of the chrominance information decision colourity amount of emphasizing of input image data;

From above-mentioned input image data, extract the colouring component extraction step of a plurality of colouring components of wanting the Accent colour degree; And

According to the colourity amount of emphasizing of above-mentioned decision above-mentioned a plurality of colouring components are carried out the colourity that colourity emphasizes and emphasize step;

Among above-mentioned a plurality of colouring components, in the colourity of predetermined colouring component is emphasized, the colourity amount of emphasizing of above-mentioned decision is adjusted.

12. an image processing program is characterized in that:

By carrying out on computers, the aforementioned calculation facility are worked as lower unit as doing:

Extract the colourity amount of the emphasizing decision unit of the chrominance information decision colourity amount of emphasizing of input image data;

From above-mentioned input image data, extract the colouring component extraction unit of a plurality of colouring components of wanting the Accent colour degree;

According to the colourity amount of emphasizing of decision above-mentioned a plurality of colouring components are carried out the colourity that colourity emphasizes and emphasize the unit; And

Among above-mentioned a plurality of colouring components, in the colourity of predetermined colouring component is emphasized, the amount of the emphasizing adjustment unit that the colourity amount of emphasizing of above-mentioned decision is adjusted.

Images